LogFAQs > #903213227

bulbinking posted...

What do you consider life to be?

Its not something we have any way to currently measure.

Sure we do. There are seven criteria that are typically employed to designate something as "living":

1. Maintains a consistent internal state (homeostasis)
2. Composed of cells
3. Carries out some form of metabolism
4. Grows
5. Responds to the environment
6. Adapts to environment
7. Reproduces

These criteria are somewhat flexible, particularly when it comes to debating whether or not things like viruses and prions qualify as living organisms. That debate mostly arises in that, intuitively, viruses and prions behave very similarly to definite living things, but they don't meet all of these criteria. If we try to apply those criteria to robots, the only one that they actually can't meet is #2, and personally, that one strikes me as being rather circular logic (something is living only if it's made of cells, which are the fundamental unit that makes up all living things) and therefore not a very good criterion to use.

bulbinking posted...

Its a very shallow and materialistic view of the world to think a physical phenomena can replace living organisms simply because they can both manipulate the environment in measurably similar ways.

Life is a physical phenomenon. Literally everything about who you are and what you do boils down to electrochemical signals in your brain and muscles. If electrochemistry can create human sentience, I see no reason why regular electricity can't create artificial sentience.

bulbinking posted...

We can create life. Its called facillitating the natural chemical processes that life uses to form and reproduce itself.

Again, you're skirting dangerously close to the naturalistic fallacy. Literally the only basis for defining something as being artificial instead of natural is if human intervention is involved in making it happen. Using that distinction as the basis for any sort of argument doesn't really do that argument any favours.

There's also the argument I've made in the past that, because it's performed by humans, human reproduction is technically artificial. Obviously, that's being a little silly, but the fact that it's a logically defensible position illustrates quite nicely how flimsy and meaningless that distinction really is. The best anyone can do to refute it is to define "natural" as "anything humans didn't have to plan," but even that becomes inadequate when you start getting into artificial intelligence.

bulbinking posted...

even if it ends up doing something it wasnt originally programmed to do (lol thats a farce)

The thing is, "what it's originally programmed to do" can be incredibly broad. To get back to the new definition I just gave for "natural," we already see artificial intelligence doing things humans never planned. Neural networks and other heuristics frequently end up generating unexpected results, and while that's technically just the expected algorithmic result of their initial parameters and the stimuli they've been given, you can say exactly the same thing about our own brains, and the path to those outcomes is too complex and specific to be reasonably predictable.

If we program a robot to think, it will think, and it will think thoughts we didn't explicitly tell it to think. It will be doing what it was programmed to do, but that doesn't mean it will behave predictably, any more so than humans doing what they're programmed to do means they will behave predictably.
---

This is my signature. It exists to keep people from skipping the last line of my posts.